REACTIONS OF THE MEXICAN STOCK EXCHANGE AGGREGATE MARKET TO LARGE DAILY PRICE SHOCKS ...catarina.udlap.mx/u_dl_a/tales/documentos/lfp/trueba... · 2018. 7. 18. · price shocks and

1

FUNDACIÓN UNIVERSIDAD DE LAS AMERICAS PUEBLA.

SCHOOL OF BUSINESS AND ECONOMICS

DEPARTMENT OF FINANCE AND ACCOUNTING

REACTIONS OF THE MEXICAN STOCK EXCHANGE

AGGREGATE MARKET TO LARGE DAILY PRICE SHOCKS: A

BEHAVIORAL FINANCE ANALYSIS

HONOURS PROGRAMME THESIS

José Trueba Patiño

ID 151464

Thesis Advisor PhD Manuela Armas Carrillo

Santa Catarina Martir, Puebla. Spring 2018

2

To my family, friends, academics and practitioners in the field.

Declaration of Authorship

I hereby declare that the thesis titled “Reactions of the Mexican Stock Exchange Aggregate

Market to Large Daily Price Shocks: A Behavioral Finance Analysis” represents my own

work in accordance with regulations. This Bachelor Thesis was realised in partial fulfilment of

the requirements for graduation from the Honours Program of the Universidad de las Americas

Puebla, UDLAP.

_______________


3

Acknowledgments

First and foremost, I want to express my deep sense of gratitude to PhD. Manuela Armas

Carrillo for her consistent and inspiring guidance, scholarly explanations and valuable

criticisms throughout the process of my work.

I am truly grateful to PhD. Felipe de Jesús Bello Gómez for his continuous support,

contribution, guidance, motivation, and high quality criticism throughout my course of work.

My special words of thanks also should go to PhD Nora Gavira Durón for her contributions

and inspiration during my learning process of both my last semester of studies and my Thesis.

A final and special mention of thanks accounts for PhD Einar Moreno Quezada whom as head

department of the Finance and Accounting School Department always guided me throughout

my Bachelor.


4

Index Abstract ...................................................................................................................................... 6 1. Introduction ........................................................................................................................ 7

1.1. Limits of the Study................................................................................................. 9 2. Literature Review ............................................................................................................. 10

2.1.1. Overview of Behavioral Finance ..................................................................... 11 2.1.2. Behavioral Finance Theories and Bias ............................................................ 12 2.1.3. From Traditional Finance Theory Towards Behavioural Finance Theory ...... 17 2.1.4. Evidence of Overreaction and Reversals ......................................................... 18 2.1.5. Evidence of Under reaction and Momentum in prices .................................... 18 2.1.6. Over and Under reaction at the same time ....................................................... 19 2.2. The Role of Behavioral Finance in The Markets and its Implications for Price Shocks 21 2.2.1. Prospect Theory ............................................................................................... 22 2.2.2. Overreaction and Contrarian Investing ............................................................ 22 2.2.3. Under reaction and Momentum Investing ....................................................... 23 2.2.4. Casual Attribution ............................................................................................ 23

3. Data and Methodology ..................................................................................................... 25 3.1. Data ...................................................................................................................... 25 3.2. Methodology ........................................................................................................ 25 3.3. Magnitude of the Positive and Negative Market Reactions Post Event .............. 28

4. Results .............................................................................................................................. 30 4.1. Single Mean Hypothesis Testing, Z test: One Sample Mean .............................. 30 4.1.1. Groping 1 ......................................................................................................... 30 4.1.2. Grouping 2 ....................................................................................................... 32 4.2. Two Sample Mean Hypothesis Testing, Z test: Two Sample Means .................. 33 4.2.1. Grouping 1 ....................................................................................................... 33 4.2.1.1. Negative Events ........................................................................................... 34 4.2.1.2. Positive Events ............................................................................................. 36 4.2.2. Grouping 2 ....................................................................................................... 37 4.2.2.1. Negative Events ........................................................................................... 38 4.2.2.2. Positive Events ............................................................................................. 40 4.3. Magnitude of the Positive and Negative Market Reactions Post Event .............. 41 4.3.1.1. Negative Events ........................................................................................... 42 4.3.1.2. Positive Events ............................................................................................. 44

5. Results Analysis ............................................................................................................... 47

5

5.1.1. Single Mean Hypothesis Testing, Z test: One Sample Mean .......................... 47 5.1.2. Two Sample Mean Hypothesis Testing, Z test: Two Sample Means .............. 47 5.1.3. Magnitude of the Positive and Negative Market Reactions Post Event .......... 49

6. Conclusions ...................................................................................................................... 51 Reference List .......................................................................................................................... 57

Index of Tables

Table 1 ..................................................................................................................................... 26 Table 2 ..................................................................................................................................... 27 Table 3 ..................................................................................................................................... 27 Table 4 ..................................................................................................................................... 28 Table 5 ..................................................................................................................................... 28 Table 6 ..................................................................................................................................... 29 Table 7 ..................................................................................................................................... 30 Table 8 ..................................................................................................................................... 30 Table 9 ..................................................................................................................................... 31 Table 10 ................................................................................................................................... 32 Table 11 ................................................................................................................................... 32 Table 12 ................................................................................................................................... 32 Table 13 ................................................................................................................................... 33 Table 14 ................................................................................................................................... 33 Table 15 ................................................................................................................................... 34 Table 16 ................................................................................................................................... 35 Table 17 ................................................................................................................................... 36 Table 18 ................................................................................................................................... 37 Table 19 ................................................................................................................................... 37 Table 20 ................................................................................................................................... 38 Table 21 ................................................................................................................................... 39 Table 22 ................................................................................................................................... 40 Table 23 ................................................................................................................................... 41 Table 24 ................................................................................................................................... 42 Table 25 ................................................................................................................................... 42 Table 26 ................................................................................................................................... 43 Table 27 ................................................................................................................................... 44 Table 28 ................................................................................................................................... 44 Table 29 ................................................................................................................................... 45 Table 30 ................................................................................................................................... 46

Index of Figures

Figure 1 .................................................................................................................................... 43 Figure 2 .................................................................................................................................... 44 Figure 3 .................................................................................................................................... 45 Figure 4 .................................................................................................................................... 46

6

"An investment in knowledge pays the best interest." Benjamin Franklin

Abstract This thesis analyses the post reactions to large daily price shocks on the Mexican capital

aggregate market using data on daily returns of its main Index (IPC) from 2009 through 2018.

It uses a variety of statistical tests and an event and post event study based on the means from

the returns of both event and post event returns. On the whole, this thesis suggests that the

market tends to both overreact and underreact to price shocks, in both cases, positive and

negative price shocks. In essence, when there is a negative price shock, the market tends to

overreact a slightly more than half of the times, while the rest of the times underreacts. While

on the other side, when there is a positive price shock, the market underreacts slightly more

than half of the times while the rest of the times overreacts. As new negative information

appears, it causes prices to significantly drop, the panic and fear among market participants

rise which is the cause of overreaction. On the other side, as new positive information is

released, it causes prices to rise, thus, investors to feel more confident on gains, and there is a

rise in greed among market participants pushing prices higher as under reaction occurs.

Theories from Behavioral Finance can potentially explain the relationship between the market

price shocks and the post event market reactions. Although the mentioned above may be true,

one possibility to explain the market anomalies are the deviations to be expected under the

Efficient Market Hypothesis. However, this thesis work is based on behavioral finance theories

which suggest potential explanation on this work.

7

“Investing is the intersection of economics and psychology.”

Seth Klarman

1. Introduction Behavioral Finance is indeed a contrast to the efficient market hypothesis, as it proposes that

the reason for market inefficiencies are the not fully rational market participants. The

traditional finance paradigm seeks to explain financial markets using models in which agents

are fully rational. Granted this view, if the Efficient Market Hypothesis holds, prices are fair

and reflect all available information, given that, there would be no investment strategy which

could earn abnormal returns.

However, it can be said that markets are not fully efficient, moreover, an idea of having

market participants whom account for the characteristics needed in order to have an efficient

market mentioned by Barberis and Thaler (2005), are characteristics which are rarely seen on

financial market participants. Certainly, new information is not instantaneously and fully

incorporated into stock prices in an unbiased fashion. Thus, prices are likely not to be fair in al

cases. In account of the research of this thesis, it can be suggested that traditional finance

theories and the efficient market hypothesis do not fully explain Financial Markets.

Hence, with the intention of achieving a better understanding of the capital market functioning,

this thesis suggests behavioral finance as a supplement and more logic explanation of the

capital market functioning and its significance for price fluctuations. There needs to be a huge

attention on market participants’ behaviour. In essence, behavioral finance replaces “perfectly

rational” economic agents from traditional finance theories for “normal” people. “Specifically,

it is the study of how psychology affects financial decisions, corporations, and the financial

markets.” (Nofsinger, 2001)

Behavioral finance emerged from the contributions of the two psychologists Daniel

Kahneman and Amos Tversky in finance field. Their work served as foundation in the field

and gave rise to the new paradigm known as “Behavioral Finance”. The rise of this field has

its roots on the difficulties faced by the traditional finance theories to explain various finance

phenomena which were better explained when the assumption of full rationality is relaxed.

Very important to realize and put emphasis on, the main purpose of behavioral finance

is not to prove any of the existence theories are obsolete, since those theories explain

successfully various scenarios to a good extent. Hence, the objective of behavioral finance

theories is to supplement the traditional theories by incorporating the cognitive psychology into

8

those theories, so thereafter, create a more complete model of the investors behaviour in the

process of decision making as market participants.

Thus, the rationale of the behavioral finance study would be that “In the field of

investments, the implications of behavioural finance are substantial, analysing investor

behaviour is an essential key in order to understand the sometimes rare and unexplainable

fluctuations of financial markets. Hence, understanding behavioural finance can provide

significant advantages for investors to the end of taking better investment decisions.”

One key factor which caused the interest to realise this thesis work, are the

psychological bias which strongly influence on market participants, whom have a strong

tendency to fail into those when taking investment choices. As human beings, there is a

“Systematic error in judgment and decision-making common to all human beings which can

be due to cognitive limitations, motivational factors, and/or adaptations to natural

environments.” Wilke A. and Mata R. (2012).

For the purpose of improving the understanding on the field, and behavioral finance

perspective on the capital market, the need to understand the main behavioral finance theories

arises. Thereafter, further on in this thesis the following theories and bias will be broadly

explained.

• Prospect Theory: Anchoring, Loss Aversion, Frame Dependence, and Mental

Accounting.

• Heuristics Theory: Misperceiving Randomness, Herding, and Overconfidence.

The mentioned behavioral theories and bias can be seen as the causes of overreaction

and under reaction in capital markets. In the same manner, contrarian investing and momentum

investing.

Although the mentioned above may be true, one possibility to explain the market

anomalies are the deviations to be expected under the Efficient Market Hypothesis (Fama

1998). However, this thesis aims to provide information suggesting that those market

anomalies are not caused by those deviations under market efficiency. A central theme on this

work is that prices sometimes overreact or underreact to some signal, moreover, often prices

over and underreact simultaneously.

As a consequence, this thesis focuses on the Mexican capital market´s short term

responses to large daily price shocks on its main index (IPC), which can be seen as an image

of the aggregate market. Price shocks in the capital market can cause investors to either gain

9

or lose wealth, which naturally causes emotional responses among market participants.

Therefore, market participants are likely to be driven by emotions such as fear and greed which

can lead to erroneous choices. Consequently, the investors behaviour in response to price

shocks will usually be a key factor on defining the direction of the post event reaction. This

thesis extends a literature review with the purpose of contributing a conclusion on the reasons

provoking overreactions or under reactions in capital markets, and if these reactions can be

explained by the aggregate behaviour of individuals.

Finally, the main hypothesis is that markets usually tend to overreact to bad news and

underreact to good news. Thus, some behavioral theories can explain the results of this market

reactions when comparing the relationship between the direction of the event and the direction

of the market responses post event. The mentioned relationship will be analysed in three

different tests further on in this thesis. This relationship has also been a debate in the field when

similar tests have used the behavioral theories to explain abnormal returns caused post price

shocks.

The already mentioned price shocks to the capital market can certainly cause market

participants to gain or lose large amounts of money, hence this thesis can also contribute to the

reader, if it is an investor, to understand the significance of their own responses, and the

responses of the aggregate market.

1.1. Limits of the Study The main weakness and advantage of this study is the fact that it is realized on the main index

of the Mexican stock exchange, representing the aggregate market. Therefore, the timeframe

post event to analyze the market reactions are just one-day and two-days window analysis, this

is due to the fact that there are various factors affecting the index, thus a longer window for the

analysis would not be accepted, since it would mean that the reactions post event are just

attributed to the price shock, which is certainly not true as there are several other reasons

causing price fluctuations in the index.

A second limitation arises out of the fact that the analysis is done on the current “bull”

market, given the points of view presented on the data and methodology chapter which explain

the reasons to choose this timeframe.

10

2. Literature Review

It has long been known by researchers in behavioural economics that the

importance of less than fully rational behaviour depends on the extent to

which rational actors can profit from the suboptimal choices of others,

especially if in the act of profiting, rational individuals push the quasi-rational

agents to behave more rationally.

Richard H. Thaler Rationale of the study. In the field of investments, the implications of behavioural finance are

substantial, analysing investor behaviour is an essential key in order to understand the

fluctuations of capital markets. Hence, understanding behavioural finance can provide

significant advantages for investors to the end of taking better investment decisions.

The traditional finance paradigm seeks to explain financial markets using models in

which agents are fully rational. Hence, agents update their beliefs correctly and make choices

for their best interests, which usually does not happen in reality. According to Barberis and

Thaler (2005), rationality means two things: First, investors update their beliefs correctly when

they receive new information, and they do so in the manner described by the Bayes’s Law.

Second, given their beliefs, investors make normatively acceptable choices, in the sense that

they are consistent with Savage’s notion of Subjective Expected Utility (SEU).

Therefore, two key paradigms within traditional finance would be enough to model

behaviour and therefore fully explain financial markets’ behaviour: (1) All new information is

available for everyone, thereupon, interpreted correctly, uniformly and instantly. Hence, (2)

Markets are efficient: new information is instantaneously and fully incorporated into stock

prices in an unbiased fashion. When the Efficient Market Hypothesis holds, prices are fair and

reflect all available information, granted so, there would be no investment strategy which can

earn abnormal returns.

"An 'efficient' market is defined as a market where there are large numbers of rational, profit-

maximizers actively competing, with each trying to predict future market values of individual

securities, and where important current information is almost freely available to all

participants. In an efficient market, competition among the many intelligent participants leads

to a situation where, at any point in time, actual prices of individual securities already reflect

the effects of information based both on events that have already occurred and on events which,

11

as of now, the market expects to take place in the future. In other words, in an efficient market

at any point in time the actual price of a security will be a good estimate of its intrinsic value."

(Fama, 1965)

Unfortunately, or maybe fortunately for some, the mentioned above does not fully

represent financial market fluctuations. Traditional Finance Theory is nowadays appealing to

be simple and incomplete. After years of research it has become very clear that traditional

finance theories do not fully explain Financial Markets.

The Efficient Market Hypothesis and Behavioural Finance theories are negatively

correlated inasmuch as one excludes the other. Traditional finance theory and behavioural

finance theory stand for contrarian thoughts that aim to explain investor behaviour. Probably

the simplest way to provide explanation of the main differences is to see traditional finance

theory as how markets would work in the ideal world. While, on the contrary, Behavioral

finance would be more accurate to provide knowledge of how markets work in the real world.

Although, having a wide understanding of both theory and reality, can help investors to better

understand fluctuations in the view of taking more optimal investment decisions.

Behavioural Finance is the approach to financial markets that argues that some financial

phenomena, meaning market anomalies, can be better explained by using models which take

into consideration investor psychology. Thus, incorporate models in which some agents are

not fully rational in order to examine what occurs when one or the two key tenets that underlie

individual rationality are soften or left apart.

2.1.1. Overview of Behavioral Finance

“Behavioral Finance is the study of the influence of psychology on the

behaviour of financial practitioners and the subsequent effect on markets”.

(Sewell, 2007) Behavioral finance aims to supplement the standard theories of finance by introducing

behavioural theories to the investment decision making. Behavioral Finance identifies the

potential reasons of stock market crashes and booms. According to Shiller (2003), the stock

crashes and booms have their roots in human errors, suggesting those errors are influenced by

behavioral finance bias. Statman (2014) says that behavioral finance replaces “perfectly

rational” economic agents from traditional finance theories for “normal” people. Statman also

suggest that the problem is not that normal people are completely irrational, but that, people

12

tend not to be completely rational, since they can be affected by some cognitive errors clarified

by behavioral finance theory.

According to Baber and Odean (1999), “Behavioral Finance relaxes the traditional

assumptions of financial economics by incorporating these observable, systematic, and very

human departures from rationality into standard models of financial markets. The tendency for

human beings to be overconfident causes the first bias in investors, and the human desire to

avoid regret prompts the second”

2.1.2. Behavioral Finance Theories and Bias Psychologists Daniel Kahneman, Paul Slovic, and Amos Tversky (1982) first introduced the

thought of psychological bias in the early 1970s. Later on, they published their findings in their

1982 book: "Judgment Under Uncertainty." They explained that psychological or cognitive

bias are the tendency to make decisions in an illogical way.

According to Wilke A. and Mata R. (2012), Cognitive bias are defined as the

“Systematic error in judgment and decision-making common to all human beings which can

be due to cognitive limitations, motivational factors, and/or adaptations to natural

environments.”

With the purpose of explaining the irrational behaviour from investors in financial

markets, behavioral economists rely on the knowledge of human cognitive behavioral theories

from psychology. Thus, this section of the thesis provides some information of the major

behavioral theories and the psychological bias that support the given theories.

Prospect Theory

Prospect Theory was first conceived by Kahneman and Tversky (1979), later on, their work

resulted in Daniel Kahneman being awarded a Nobel Prize for Economics. The theory aims to

explain the irrational human behaviour when assessing risk under uncertainty. In essence, it

proposes that investors are not persistently risk averse; instead they are risk averse in gains but

risk takers in losses.

The main idea of prospect theory is that people tend to take too much into consideration

their changes in wealth, instead of just their comprehensive level of wealth (Jordan et al. 2015).

13

This theory can be seen in four main psychological bias: Anchoring, loss aversion, frame

dependence, and mental accounting.

• Anchoring tends to happen when agents fix a certain reference point for investments.

Investor will receive utility from gains and loses due to that reference point. According

to Duxbury (2015) that reference point is usually the share price. Hence, the problem

occurs as the investor will be putting too much focus on the given reference point

instead of taking into consideration the level of his wealth. This can be explained when

an investor who has held a winner stock for a long time in relation to its purchase price

will refuse to sell the given stock. Thus, the stock’s price could have a substantial factor

on the decision making process of whether selling or keeping the stock (Kliger et al.

2014). Hirshleifer (2015) suggested that market participants should focus more on

where the stock value lies in reference to its current covariance with their portfolios.

• Loss Aversion. As a matter of fact, people hate losing. Actually people even hate losing

more than they enjoy winning. Thaler (1999) said that losses negatively affect investors

more or less as twice as wins positively affects them. Moreover, media encourages this

loss aversion by causing panic on financial news when the market is going down

(Doviak, 2016). Basically, loss aversion occurs when invertors are reluctant to sell any

stock that could result in a loss.

• Frame dependence is a theory which suggests that depending how a problem is

described will affect how an individual will react to it. Some frame dependence may be

caused by highlighting a reference point or simple approaching a decision while

isolating the factor which is related to (Hirshleifer, 2015). Frame dependence is strongly

related to anchoring. A good example of the already mentioned above would be when

an invertor looks at a stock`s performance in reference to its purchase price instead of

taking in account the diversification it adds to his portfolio.

• Mental Accounting. According to Jordan et al. (2015), mental accounting is the

tendency to separate money into “mental buckets” resulting in a different approach of

value and risk for each of the buckets differently. The most common mistake among

market participants is not to threat money as fungible, meaning “freely exchangeable

by another of like nature”. Hence, investors tend to fail in keeping a general view of all

of their assets and desired outcomes together.

Thaler (1999) displayed another example of mental accounting with the “house money

effect”, which basically proves that when gambler is winning, thus “playing with house

14

money”, they tend to become risk seeking with their money. Likewise, investors who

are having high returns tend to become much less risk averse in their investments

(Thaler, 1999).

The explanation for the mentioned above is because individuals usually threat money

differently depending on how it was earned or lost.

Heuristics Theory

“Heuristics are simple efficient rules of the thumb which have been proposed to

explain how people make decisions, come to judgments and solve problems,

typically when facing complex problems or incomplete information. These rules

work well under most circumstances, but in certain cases it leads to systematic

cognitive biases”

Daniel Kahneman (Parikh, 2011). Heuristics is a problem solving method based on shortcuts in order to simplify the decision

making process. Thus, heuristics can provide good enough solutions for taking choices when

there is a limited time frame for the decision making process.

According to Greenwood and Nagel (2009), prior experience affects investors’

decision, and oftentimes this prior experience leads to heuristic investment strategies. Investors

can also rely too much on certain habits without giving them enough thought (Hirshleifer,

2015). Investors usually tend to take into consideration the short sighted belief that “the

majority is always right”, which also leads them to fall into herding (Andersson, Hedesstrom,

and Garling 2014). Choi, Laibson, and Madrian (2009), say that investors could over

extrapolate their own prior investment performance when making decisions.

Thus, everyone could well fall into heuristics when making investment decisions, and

they can do so unintentionally. Therefore, investors are strongly suggested to become aware of

their own heuristics so as to make more accurate investment decisions.

15

Misperceiving Randomness

“Thus, the representativeness heuristic can encourage people to expect

intuitively past price changes to continue, even if they know, from professional

training, that they should not expect this.”

Robert Shiller Misperceiving Randomness is also known as the representative heuristics, Kahneman and

Tversky (1979) were the pioneers to define it. They defined representative heuristics as the

action of people trying to predict future outcomes based off random past events. Shiller (2003)

suggested that human minds can be seen as a patter seeking device. When investors are trying

to estimate future outcomes based off technical analysis (past data and/or graphs), the human

mind tends to fall for the actual poor probability that the current pattern will continue (Shiller,

2003).

Market participants tend to have a combination of recency bias, which is defined as

putting too much attention on recent events when forecasting future outcomes (Sinha, 2015),

and use the law of small numbers. This occurs as people rely too much on small samples to

draw conclusions (Tversky and Kahneman, 1971).

Herding

“Most people get interested in stocks when everyone else is. The time to get

interested is when no one else is. You can’t buy what is popular and do well.”

Warren Buffett

“By the time any view becomes a majority view, it is no longer the best view:

somebody will already have advanced beyond the point which the majority have

reached.”

Friedrich Hayek The herding bias is among the most common mistake within the investors’ community as

usually they tend to follow the investment decision of the majority. According to Jordan et al.

(2015) investors constantly fall into herding by being subsequent to do what the majority are

doing with their investments. The fundamental reasons market participants fall into herding is

because they believe that their own information is incomplete or has a lack of quality,

16

henceforth, they assume other investors have higher quality information (Sinha, 2015). Herding

also tends to happen due to the lack of information at all.

Roider and Voskort (2016) tested the theory of herding among a group of investors and

successfully proved it. There are tons of evidence suggesting that analysts are strongly

persuaded by others as they fix their forecasts to follow the herd. One of the main reasons of

herding is the fear to fail on their own.

Herding causes several issues to capital markets by destabilizing prices and causing

bubbles (Andersson et al. 2014). This occurs as investors make decisions based off of the

majority’s decisions which have been proved not to be right all the time. Thus, herding can

cause volatility and strong movements in prices if the herd changes beliefs or confidence in a

given stock or the aggregate market. Conforming to Venezia et al. (2011), herding is

significantly positive correlated to market volatility in capital markets. In conclusion, this could

create a never-ending cycle as investors panic while falling into herding and therefore

potentially increase market volatility.

Overconfidence

"It's not how right or how wrong you are that matters but how much $ you make

when right and how much you do not lose when wrong"

George Soros One of the most common mistakes among the investors community is the overconfidence

nature in relation to their own investment strategies and market estimates. According to Thaler

and Barberis (2002), the events that people believe will occur 98% of the times only happen

60% of the times. Moreover, events that people are certain that will not occur actually happen

20% of the times (Thaler and Barberis, 2002). Therefore, the mentioned overconfidence among

investor may lead them to take inappropriate decisions.

"In this business if you're good, you're right six times out of ten. You're never

going to be right nine times out of ten"

Peter Lynch Another sing of overconfidence is when an investor is familiar with a stock and

therefore the perception of risk decreases. Investors are more likely to bet on financial

instruments they are familiar with instead of taking a better gamble in something unfamiliar

17

(Heath and Tversky, 1991). Also, overconfidence can lead to affect diversification as investors

tend to place larger investments on individual securities they feel more familiar with

(Hirshleifer, 2015). The familiarity with a stock has no relation to its performance, thus

investors should see all financial instruments as equal.

2.1.3. From Traditional Finance Theory Towards Behavioural Finance Theory Kahneman and Tversky (1979) were the first to challenge the EMH with their blockbuster

report on prospect theory. Their work proposed Prospect Theory as an alternative to the

Expected Utility Theory. They found that investors tend to weight alternatives wrong in

relation to dealing with risk.

Later on, Thaler (1980) contributed to Behavioral Finance by arguing the prevalent

economic theory from that time of how consumers made their purchasing decisions. This

preponderating theory was the rational maximizing model, which stated that consumers use all

available information for any purchasing decisions, hence they make the most rational decision.

Thaler (1980) found many decision making errors people tend to do, mistakes such as

underweighting opportunity costs, regret aversion, and fail to ignore sunk costs.

De Bondt and Thaler (1985) realized a study of how individual’s cognitive bias could

produce mispricing of equities in the New York Stock Exchange to prove overreactions in the

stock market. Their study proved that people in fact overreact to bad news. This study is

mention with further detail in the section of Evidence of Overreaction.

Andreassen and Kraus (1988) also challenged the EMH by proving that when investors

analyse historical prices, they tended to behave as if they extrapolate past price changes when

the prices appear to exhibit a trend relative to period-to-period variability. Hence, they let that

have an impact on their decision making in their investments. This also suggests, that bubbles

can be generated if agents are preconditioned by past experience to form expectations of

bubbles.

Shiller (2003) stated that academic discussion began to shift towards an analysis of

human psychology role in the financial markets’ fluctuations in the 1990s. Thaler (1999)

applied behavioural finance theory to successfully predict the crash of the Internet stock boom,

he blamed the irrational investor to be the cause of the bubble. Finally, behavioural finance is

nowadays clearly a legitimate tool to be used in investment theory.

18

2.1.4. Evidence of Overreaction and Reversals

"The contrary investor is every human when he resigns momentarily from the

herd and thinks for himself"

Archibald MacLeish DeBondt and Thaler (1985) split stocks by their previous returns over the past 3 years, they

separated stocks in winners and losers depending on their previous returns. In the mentioned

analysis, they find that over a long term, stocks that had performed poorly (losers) go on to

outperform stocks that had performed very well (winners). Therefore, it could be said that

investors’ overreactions are the causes of the already mentioned above. Also, this would

suggest evidence of Contrarian Investment. The reason for this is because winner stocks are

over bought (Above fair value) and loser stocks are oversold (under fair value), hence profits

occur as winner stocks retreat to their fair value and loser stocks rally to their fair value.

As well as, there are studies that show evidence of overreactions when analysing shorter

periods of time. Zarowin (1989) realises an analysis as the mentioned above by Debondt and

Thaler, but he does it by analysing the performance of stocks over the previous month and then

examines returns over the next month. Once again, he finds that previous losers outperform

previous winners. Bremer and Sweeney (1991) state that after a ten percent daily decline in

prices, stocks perform a reversal to initial prices by having positive abnormal returns for the

next two days.

2.1.5. Evidence of Under reaction and Momentum in prices As well as Overreaction, empirical studies have also found examples of under reaction. Under

reaction causes momentum in prices as stock prices keep on moving in the same direction as

the information is slowly incorporated into stock prices. Momentum could be seen as the first

stage in the short term market’s reactions to price changes. Jegadeesh and Titman (1993) build

portfolios just as Debondt and Thaler but used the prior six months returns and examined the

reaction for the following six months return. Jegadeesh and Titman found evidence of

momentum in the short term, meaning prior winners outperform and prior losers underperform.

Since momentum has been proved to exist in the short term, it could be seen also as the first

stage of overreaction. As it is mentioned above, winner stocks are over bought (Above fair

value) and loser stocks are oversold (under fair value). This causes Momentum in prices, hence

19

it could be seen as the period of time where investors cause a tendency in prices due to short

term under reaction.

Underreaction and price drifts are usually related to important public announcements.

Therefore, positive news can lead to subsequent positive abnormal returns, and negative news

lead to subsequent negative abnormal returns. Some examples of negative events that could

lead to negative abnormal returns include stock splits (Grinblatt, Masulis, and Titman 1984),

open market repurchases (Ikenberry, Lakonishok, and Vermaelen 1995), public

announcements of previous insider trades, analyst recommendations, among some others.

However, the most famous and most well known and studied market anomaly is the Post

Earnings Announcement Drift (PEAD). PEAD is the tendency of a stock’s cumulative

abnormal returns to drift in the same direction of the earnings surprise (positive or negative)

for a considerable period of time. Brown (1997) states that the drift following standardized

unexpected earnings is very significant, hence it can also lead to momentum in prices.

According to Kim (2002), the firms which consistently achieve or beat analysts’

expectation earn a market premium in proportion to the times the firms have met their target.

Anyhow, the mentioned premium is lost when firms miss their target earnings, suffering a

proportionate reversal. Rapidly appreciating growth stocks tend to outperform during a boom,

but also they tend to underperform in the long term if they do not meet expectations.

Empirical studies from the earnings surprise propose that investors use heuristics as

they behave by the rule of the thumb, meaning the future will represent the past. Consequently,

this causes momentum in prices while the market responds and incorporates information into

prices. Finally, prices tend to continue moving in the same direction for a period of momentum

while investors fully incorporate the event´s information.

2.1.6. Over and Under reaction at the same time One possibility to explain the market anomalies already mentioned are the deviations to be

expected under the Efficient Market Hypothesis (Fama 1998). However, this thesis aims to

provide information suggesting that those market anomalies are not a caused by those

deviations under market efficiency. A central theme on this work is that prices sometimes

overreact or underreact to some signal, moreover, often prices over and underreact

simultaneously.

Daniel, Hirshleifer, and Subrahmanyan propose a model in which overconfidence

causes overreaction, and self attribution leads to underreaction. “Stock prices tend to overreact

20

to private information signals and underreact to public signals, thus positive return

autocorrelations can be a result of continuing overreaction.” (Kent, Hirshleifer, and

Subramanian 1998).

According to Porterba and Summers (1998), there is a transitory part of prices which

leads to positive autocorrelation over short terms, but negative autocorrelation over long terms.

A Non-Random Walk Down Wall Street by Andrew W. Lo and Craig Mackinlay is compound

by eleven papers which all come to the same overall conclusion that there is positive

autocorrelation in stock prices over short terms, but negative autocorrelation over long terms.

Also, Barberis, Shleifer, and Vishny (1998), suggest a model which integrates

anchoring and heuristics where conservatism first causes underreaction but trend chasing leads

to overreaction in the long term. By the same token, Hong and Stein (1999) propose a model

of over and under reaction suggesting that fundamental analysis is the reason of underreaction

while technical analysis leads to overreaction.

In brief, the mentioned above suggests that there is momentum in the short run but mean

reversion in the long term. Stock prices tend to first drift (sign of underreaction), but thereafter

prices correct (sign of overreaction).

Moreover, Kaestner (2006) suggests that the market whether over or underreact to an

earnings surprise depends on how surprising is the event. He finds that the level of surprise is

what defines the level of over or underreaction. Thus, it can be said that the greater the surprise

from the price announcement, the stronger the market reaction.

Although this behavioural models may be true, traditional finance models also suggest

explanation to those movements in prices dismissing them as methodological or statistical

flaws. Some authors of traditional finance models believe that changes in risk that are

compensated by changes in returns appear to be abnormal. As an example, it could be argued

that prior losers are riskier than prior winners, hence they explain momentum as a

compensation for risk in a given period of time.

“From a simplistic point of view, consider a bull market in which stock prices

have been rising. Investors see their portfolio values increasing. Those who

have been sitting on the side-lines hear how their friends have made money in

the stock market. Not wanting to miss out on these returns, they join in. The

average investor is hopeful and confident that the trend of rising stock prices

will continue. Of course, as these investors place more and more money in the

21

market, stock prices do rise; in economic jargon, as the quantity of investors in

the marketplace increases, the demand for stocks increases, driving stock prices

even higher. The optimistic view of the market participants drives prices even

higher. Seeing that they were correct, investors become overconfident and

greedy and purchase even more stocks. At the peak of optimism, investors have

placed all their available money in the stock market. At this point, there is no

new money coming into the market to continue fuelling the increase in demand

that has been driving price upward. Thus, there is no more fuel to keep stock

prices rising, and the stock market reaches a peak.”

“Conversely, when investors are pessimistic and fearful, they begin to sell

stock. As the level of pessimism rises in the market and more investors sell,

stock prices fall. These falling prices lead more and more investors to feel

fearful and decide to sell their shares. When investors are the most pessimistic

and fearful, they have withdrawn all of their money out of the market. The

downtrend that has been fuelled by investors leaving the markets ends, and the

market reaches a bottom.”

Charles Kirkpatrick and Julie Dahlquist

2.2. The Role of Behavioral Finance in The Markets and its Implications for Price Shocks

“One of the funny things about the stock market is that every time one person buys, another

sells, and both think they are astute”

William Feather

The Decision making process is a various steps process which implicates the analysis of diverse

factors influencing a given problem. Probably taking investment decisions is one of the most

crucial decisions for everyone. Optimal decision making for investing in the stock market

requires a deep understanding of investors rationality and more important, irrationality in a

global perspective. Given the past market crashes, there is been a lot of attention on studying

irrational investor behaviour. “Behavioral Finance is becoming an integral part of decision-

making process because it heavily influences the investors’ performance” (Banerjee, 2011).

22

This section aims to explain the role of behavioral finance in capital markets, in the

foreground of market reactions following significant price shocks. Thus, this part is split into

the most inclusive factors causing the reactions following a price shock.

2.2.1. Prospect Theory It could be said that if investors behave as explained in prospect theory, they would react

differently to positive and negative price shocks. This behavioral theory suggests that following

significant gains, such as a strong positive price shock, investors would become more risk

averse, thereafter, they would sell in order to obtain profits. For the cases of analysis provided

further on in this thesis, this would imply a negative mean of the returns from the following

days after positive events.

On the contrary, this theory implies that investors tend to become more risk seeking

following losses. Thus, instead of selling after a negative price shock, investors would hold the

stock or even buy more with the purpose of gambling back their losses (Gambler’s Fallacy).

Therefore, since market participants tend to become more risk seeking after looses and improve

their confidence on the market, this would manifest itself in the analyses provided further on

in this thesis as a very poor market reaction following a negative event, resulting in a very small

mean, either a positive or a negative mean close to 0 of the returns from the following days

after negative events.

2.2.2. Overreaction and Contrarian Investing Overreaction would imply similar results as in Prospect theory in terms of the tests provided

further on in this thesis. Overreaction in the market would imply the initial price shock to be

too strong mainly caused by emotions on the day of a price shock. As a result, prices would

drift towards unfair value on the event day, as well as in the same direction of the event,

meaning overvalued for positive events or undervalued for negative events. According to

Atkins and Dyl (1990), whom examined the behaviour of common stock prices following a

large change in prices in single days of trading, they found evidence suggesting that the stock

market appears to overreact, especially following a negative price shock. In accordance to their

study, prices should fall following positive events and raise after negative events as to return

to fair value. Thus, returns of the days of the events would be negatively correlated with the

returns of the days following the event. In the cases of analysis in this thesis, this would suggest

23

a positive mean of the returns from the following days of the events in the case for negative

events, likewise a negative mean for the case of positive events.

Finally, if there is overreaction in the market this would suggest prove and efficiency

of contrarian investing strategies. This would happen as stocks that had performed poorly

(losers) go on to outperform stocks that had performed very well (winners). The reason for this

is because winner stocks are over bought (Above fair value) and loser stocks are oversold

(under fair value), hence profits occur as winner stocks retreat to their fair value and loser

stocks rally to their fair value.

2.2.3. Under reaction and Momentum Investing Under reaction mainly implies a price drift following a price shock in the same direction of the

price shock. The magnitude and length of the period of time for the price drift would be mainly

attributed to the magnitude on the news’ information impact which causes the original price

shock. At the same time, this would cause momentum following price shocks. According to

Bernard and Thomas (1990), prices would continue to increase following positive price shocks,

as well as, prices would continue to fall after negative price shocks due to too moderate initial

reactions in events.

2.2.4. Casual Attribution Casual attribution could be seen as the logic-sequence explanation of both under and over

reaction. This theory explains the market reaction in relationship to whether the investors

attribute the price shock to a stable or non- stable cause. Given the prior day’s return can

significantly impact investor’s expectation on the market, but those expectations did not

account during the prior day shock, drifts or reversals in prices post event might be attributed

to whether the price shock had a stable or non-stable cause.

Since investors usually use hindsight (understanding of a situation or event only after

it has happened or developed), and tend to fail into herding when a significant price shock

occurs, they usually do not absorb information and their future expectation on the day of the

event.

Henceforth, if investors attribute the cause of the event is stable (solid fundamentals

and technical analysis supports the cause), they would believe the past results will repeat. As a

consequence, positive events would lead investors to expect more gains and increase optimism,

24

while on the other side, negative events would lead investors to expect more losses and increase

pessimism. Taking this into consideration, it could be said that the behaviour would be similar

to under reaction, prices would continue to increase following positive price shocks, as well as,

prices would continue to fall after negative price shocks.

On the other hand, when investors believe the cause of the event is non-stable, the same

behaviour of overreaction would happen, there would be a reversion to the mean pattern. This

would imply positive market reactions following negative events, and negative reactions

following positive events. This tends to occurs because investors usually believe that non-stable

causes do not repeat. Since they look back in time and do not expect the extreme results of the

prior events will happen again. This may also have relationship to whether the cause of the

event had solid fundamentals or not. Thus, this could be seen as the logic in the market reactions

post an overreaction.

25

3. Data and Methodology In this chapter, the data and methodology used in the thesis for analysis, and hypothesis testing

will be explained in detail.

3.1. Data The data used for this Thesis analysis are the closing prices and returns of the S&P/BMV IPC

(Índice de Precios y Cotizaciones). The time frame to analyse market reactions is from 2009 to

March 2018, closing prices were retrieved from the S&P Dow Jones Indices official site. The

IPC is closely followed by investors, as well as, it is very well known to be the best composite

index to represent the Mexican aggregate capital market. It is the best single gauge of Mexican

equities. Thus, daily changes within the IPC are widely recognized by market participants

causing some vast different reactions to changes over significant daily percentage changes.

The reasons for choosing this period of time for the analysis are: firstly, to exclude the

events near to the 2008 crash, which produced returns so extreme that would strongly affect

the results for this post-event analyses. Secondly, it is certainly known that before the 2008

crash there were far less regulations in stock markets. Thus, the market from before 2008 does

not represent the same context as the market from the last few years after the crash. Hence, this

analysis aims to provide results which represent the current context of the actual capital market.

3.2. Methodology Events in daily price changes are defined in a similar manner as Lowe (2008), who also defined

events similarly to Bremer and Sweeny (1991), Cox and Peterson (1994), Larson and Madura

(2002), and Sturm (2003). The mentioned above besides Lowe, analysed single securities for

which they used price shocks of 10% to individual securities to define an event, however, this

criterion can not be used for an aggregate index which is by far less volatile. By the same token,

Lowe (2008) classified events as a daily percent change greater than 2% or less than -2%. Also,

with the purpose of analysing mild events, he uses a 3% edge to define positive and negative

events and examine the 1.5% to 3% range to which he calls mild events.

26

This thesis defines events in a similar same manner as Lowe (2008), but with a slide

change on the event definitions for the return percentages. There are two different groupings

for the event definitions. The Grouping 1 and Groping 2 are classified as following in Table 1:

Table 1

Thus the events showed above are defined in relation to the daily return on the IPC.

The daily raw return is calculated as in the following formula (equation 1):

(1)

This thesis analysis the post-event market returns based on the given event in order to

asses the market reaction to significant daily price changes. To the end of analysing the market

behaviour following the event, two post event windows are used. This criterion is also used in

Sturm (2003), Chen and Siems (2002), and Lowe (2008) in a similar manner. Thus, the post

event windows are in the following manner:

v The IPC raw return on the following day of the event (RR1)

v The average IPC raw returns (ARR2) for the following 2 days of the event.

If events are defined to take place at t=0, then the RR1 is simply the raw return on the

following day, and ARR2 is obtained as in the following formula (equation 2):

2

ARR2t= 1/2* ∑RRt (2) t=1

Grouping 1

Groping 2

-1.5%<NNE<0% 0%<PNE<1.5% 1.5%<=PE

Event Definitions

Negative Event Negative No Event Positive No Event Positive Event

Negative Event Negative No Event Positive No Event Positive Event Events

Prior Day Returns

Events

Prior Day Returns NEG<=-1.2%

NEG<=-1.5%

-1.2%<NNE<0% 0%<PNE<1.2% 1.2%<=PE

IPC t+1 -IPC t=0

IPC t=0

Daily Return t+1 =

27

The raw return on the following day of the event (RR1) and the average raw returns on

the following 2 days (ARR2) are taken for reference to asses the market reaction to the events

already mentioned above. The market performance following the events is statistically

analysed in order to gauge whether there is any evidence suggesting Under reaction for the

Positive Events or Overreaction for the negative events. Hence, under reaction would suggest

momentum investment, and on the other hand overreaction would suggest contrarian

investment.

All the statistical tests analysis done in this thesis are realised with the following alpha

values (Table 2):

Table 2

Firstly, a single mean analysis is conducted in order to asses if the post event reactions

are positive or negative. However, in order to gauge the statistical significance, a Z test for one

sample mean hypothesis testing is conducted as following (Table 3):

Table 3

Secondly, a Z test for two sample means is done in order to compare the Negative

Events’ post event reaction mean against the Negative No Events’ post event reaction mean.

Also, the test compares the Positive Events’ post reaction mean against the Positive No Events’

post reaction mean.

The mentioned analysis has the purpose of assessing the main Alternative Hypothesis’

of this thesis, which are described in the following table (4):

Lower Tailed Upper TailedZ<-Z⍺ Z>Z⍺

⍺=0.05 Z<-1.645 Z>1.645 Z<-1.96 Z>1.96⍺=0.1 Z<-1.280 Z>1.280 Z<-1.645 Z>1.645

Z<-Z⍺/2 or Z>Z⍺/2 Two-Tailed

Alternative Hypothesis (Ha)Rejection Regions to the Null Hypothesis for the given ⍺ (alfa) Values

Event Definition Ivestor Type Causal Attibution Market Reaction at t=0Emotions Prospect Theory

Contrarian Non-stable Overreaction Fear, panic AppliesMomentum Stable Underreaction Greed N/A

Momentum Stable Underreaction Greed N/AContrarian Non-Stable Overreaction Fear Applies

Positive Event (PE) x̅ PE=0x̅ PE>0x̅ PE<0

Single Mean Hypothesis Testing

Test Statistic: Z test: One Sample Mean Behavioral Finance Suggested Cause for market

Null Hypothesis (Hn) Alternative Hypothesis (Ha)

Negative Event (NE) x̅ NE=0x̅ NE>0x̅ NE<0

28

Table 4

This thesis aims to find evidence on Overreaction for Negative Events and Under

reaction for positive events. Thus, in the same manner, Contrarian Investing following

Negative Events, and Momentum Investing following Positive Events.

Therefore, in order to support the mentioned hypothesis, the RR1 and ARR2 means

from Negative Events must be statistically higher than the same means from the Negative No

Events when tested in the Z test. Thus, if results are as mentioned, it would suggest there is

more overreaction for big falls on the index than with small drops of the IPC.

On the other hand, in order to support the under reaction hypothesis for Positive Events,

the RR1 and ARR2 means from Positive Events must be statistically higher than the means

from the Positive No Events when tested in the Z test. Thus, if results are as mentioned it would

suggest there is more under reaction, and therefore momentum for big price drifts upwards on

the index than with small drifts upwards of the IPC.

3.3. Magnitude of the Positive and Negative Market Reactions Post Event A third analysis is done with the purpose of assessing the market reactions post Negative and

Positive Events. Thus, this analysis is realized for a third grouping of data (Grouping 3), which

is classified as the following table:

Table 5

As it can be seen in the table above (5), grouping 3 is very similar to grouping 1 in

terms of the range for classifying events. However, groping 3 does not take into consideration

the days between the negative and positive events, thus the range from -1.2% to 1.2% is not

taken into consideration for this model. The reason for the already mention is because the

Event Definition Ivestor Type Causal Attibution

Market Reaction Emotions Prospect Theory



Null Hypothesis (Hn)

Test Statistic: Z test: Two Sample for Means

Hypothesis Testing

Negative Event (NE)

Positive Event (PE)

x̅ NE>x̅ NNEx̅ NE<x̅ NNE

x̅ PE>x̅ PNEx̅ PE<x̅ PNE

x̅ NE=x̅ NNE

x̅ PE=x̅ PNE

Alternative Hypothesis (Ha)

Behavioral Finance Suggested Cause for Market Post Event Reactions

Event Definitions for Groping 3

NEG<=-1.2% 1.2%<=PE

Event Definitions Negative Event Positive Event

Prior Day Returns

29

purpose of this model is to asses the market reactions post event, hence, the range between

events is considered as non significant events. In conclusion, this groping is focused merely on

price shocks, both positive and negative, thereafter, the post event market reactions can be

judged in terms of their relationship to behavioral finance theories.

The post event window to analyse will be just the following day after the event. In that

case, the post event window is in the following manner:

• The IPC raw return on the following day of the event (RR1)

Once defined the returns to analyse, the post event given window analysis for returns

is split into positive and negative reactions, meaning the positive those returns above 0, and

negative those below 0. Hence, the returns to analyse would be in the following manner

(Table6):

Table 6

Thereupon, with the intention of analysing if the mentioned market reactions suggest

any evidence of over and/or under reaction for the given events a magnitude analysis of means

is done. This analysis consists in comparing the means of the positive and negative post event

reactions in comparison to the mean of the Event. Consequently, if there is a significant price

drift both in magnitude of means and economically significant in the same direction of the

event (Ex: positive for positive events), then it would suggest evidence of underreaction. On

the other hand, if it occurs a mean reversal, which would mean a price drift on the opposite

direction of the event, same as the case before, it would need to be significant both in magnitude

of means and economically, this would suggest evidence of overreaction.

Finally, a set of graphics is shown in order to provide a wider panorama from a visual

perspective to asses the magnitude for post event reactions. First, in each graphic, the mean of

the whole IPC returns for the period of analysis is shown for reference. Second, the graphics

illustrate the events’ mean followed by the post event mean (one graphic for negative post

event and one for positive post event).

% Mean n % Mean nPositive,Reactions %RR1>0 51.282% 0.8549% 100 55.224% 1.0358% 111Negative,Reactions %RR1<0 48.718% =0.9804% 95 44.776% =0.7964% 90

Data from Groping 3

NEG<=-1.2% 1.2%<=PE


Number of Events

Market,Reactions,Mean of the Event !1.9391% 1.9155%

195 201Prior Day Returns

30

4. Results In this chapter the results from the statistical analysis will be provided with a short

interpretation and comments of each statistical test. First, the results from the single mean

hypothesis testing will be shown in order to asses whether the post event reaction is positive or

negative. Second, the two sample mean hypothesis testings are provided with the purpose of

gauging whether there is supporting evidence on Overreaction for Negative Events and

Underreaction for positive events when comparing the means of the Negative Events against

the means of the Negative No Events; on the other hand, the means from the Positive Events

against the means from the Positive No Events. Thus, likewise, Contrarian Investing following

Negative Events, and Momentum Investing following Positive Events. Finally, a third analysis

is provided with the purpose of assessing the direction of the market reactions post Negative

and Positive Events in a different grouping of data. The purpose of this last model is to asses

the market reactions post event, focusing merely on price shocks, both positive and negative,

thereafter, the post event market reactions can be judged in terms of their relationship to

behavioral finance theories.

4.1. Single Mean Hypothesis Testing, Z test: One Sample Mean Table 7

Note: For the following Tables the “Non-Rejection” Region means accepting the Null

Hypothesis; therefore, the “Rejection” region refers to reject the reject the null hypothesis and

accept the alternative hypothesis.

4.1.1. Groping 1 Table 8

Event Definition Ivestor Type Causal Attibution Market Reaction at t=0Emotions Prospect Theory



Positive Event (PE) x̅ PE=0x̅ PE>0x̅ PE<0

Single Mean Hypothesis Testing

Test Statistic: Z test: One Sample Mean Behavioral Finance Suggested Cause for market

Null Hypothesis (Hn) Alternative Hypothesis (Ha)

Negative Event (NE) x̅ NE=0x̅ NE>0x̅ NE<0

Number of Events Total= 2321 195 8.402% 907 39.078% 1018 43.860% 201 8.660%Prior Day Returns

Event Definitions Negative Event Negative No Event Positive No Event Positive Event

Results for Groping 1

NEG<=-1.2% -1.2%<NNE<0% 0%<PNE<1.2% 1.2%<=PE

31

Table 9

As it can be seen above, results from the Z test do no vary between the post event one

day and the post event two days. In both cases the means from the negative events and negative

no events are not indistinguishable from cero. Hence, this statistical test would suggest that

there is no supporting evidence of significant strong market reactions following negative

events. On the other hand, in the case for the Positive Events, the statistical test proves that

there is a positive market reaction. (Table 9)

Additionally, for both cases, one and two days’ analysis for the positive events and the

positive no events the results show that the means are indistinguishable from cero; more over,

the means are statistically higher than cero in all cases. This result would suggest there is a

positive market reaction following positive events. However, this is not enough evidence for

suggesting either underreaction nor momentum following a positive price shock.

Mean Z scoreZ test Lower Tailed Upper Tailed Two- Tailed Lower Tailed Upper Tailed Two- Tailed ⍺=0.05 Results Non-Rejection Non-Rejection Non-Rejection Non-Rejection Non-Rejection Non-Rejection⍺=0.1 Results Non-Rejection Non-Rejection Non-Rejection Non-Rejection Non-Rejection Non-Rejection

Mean Z scoreZ test Lower Tailed Upper Tailed Two- Tailed Lower Tailed Upper Tailed Two- Tailed ⍺=0.05 Results Non-Rejection Rejection Rejection Non-Rejection Rejection Rejection⍺=0.1 Results Non-Rejection Rejection Rejection Non-Rejection Rejection Rejection

-0.875286124-0.03923%

2.374866224 2.506956614

-0.02892%

0.21539% 0.07106%

Hypothesis Testing for Post Event One Day Raw Returns Reactions Negative Event Negative No Event

Positive Event Positive No Event

-0.452840845



Negative Event Negative No Event

Positive Event Positive No Event

0.04783% -0.00114%

2.038399410.15186% 0.04262%

2.217375796

0.76392435 -0.048185621

Hypothesis Testing for Post Event Two Day Raw Returns Reactions

32

4.1.2. Grouping 2 This section displays the results of the Grouping 2 for the single mean hypothesis testing. First,

in the following table (10) comes the general information and data for grouping 2

Table 10

Table 11

In the case for Grouping 2, the Z tests’ results are the same as in for Grouping 1. Thus,

short interpretation of the results can be found in t section 5.1.1.

Table 12

In the case for Grouping 2, the Z tests’ results are the same as in for Grouping 1. Thus,

short interpretation of the results can be found in t section 5.1.1.

Number of Events Total= 2321 124 5.343% 978 42.137% 1097 47.264% 122 5.256%



Prior Day Returns NEG<=-1.5% -1.5%<NNE<0% 0%<PNE<1.5% 1.5%<=PE



Negative Event Negative No Event

Positive Event Positive No Event 0.23978%

2.099294080.07874%

2.771417029

-0.01340%-0.119913482

-0.03295%-1.023323741

Hypothesis Testing for Post Event One Day Raw Returns Reactions



Negative Event Negative No Event 0.04990%

0.5965076350.00216%

0.094579706

Positive No Event 0.17270%

2.0544600470.04817%

2.278937475

Hypothesis Testing for Post Event Two Day Raw Returns Reactions

Positive Event

33

4.2. Two Sample Mean Hypothesis Testing, Z test: Two Sample Means This part shows the results from the two sample mean hypotheses testing. (table 13) Table 13

The following Z tests results have the purpose of assessing the means of the “events”

in relationship with the means of the “no events”. Therefore, the Negative Events’ post event

reaction mean against the Negative No Events’ post event reaction mean. Also, the test

compares the Positive Events’ post reaction mean against the Positive No Events’ post reaction

mean. The results’ interpretation is based on the table of hypothesis testing above. However,

even if the statistical results support the alternative hypothesis’, there needs to be use of criteria

in terms of means’ magnitude and economic significance for real life model applications.

4.2.1. Grouping 1 The following table (14) provides the general information and date of Grouping 1 Table 14

Event Definition Ivestor Type Causal Attibution

Market Reaction Emotions Prospect Theory



Null Hypothesis (Hn)

Test Statistic: Z test: Two Sample for Means

Hypothesis Testing

Negative Event (NE)

Positive Event (PE)

x̅ NE>x̅ NNEx̅ NE<x̅ NNE

x̅ PE>x̅ PNEx̅ PE<x̅ PNE

x̅ NE=x̅ NNE

x̅ PE=x̅ PNE

Alternative Hypothesis (Ha)

Behavioral Finance Suggested Cause for Market Post Event Reactions

Number of Events Total= 2321 195 8.402% 907 39.078% 1018 43.860% 201 8.660%Prior Day Returns



NEG<=-1.2% -1.2%<NNE<0% 0%<PNE<1.2% 1.2%<=PE

34

4.2.1.1. Negative Events Table 15

In the case for the Negative Events’ Z test analysis for a given alpha of 5%, the results

above demonstrate that the difference between the mean of the Negative One Day and the

Negative No Event One Day is not indistinguishable from 0. The already mentioned is

confirmed in the P-Value for one tail and the obtained z. (Table 15)

As well as, in the case for the Negative CRR 2 Days in comparison with the Negative

No Event CRR 2 Days, the result is the same: the difference in means is not indistinguishable

from 0.

⍺=0.05z-Test: Two Sample for Means

Negative One Day Negative No Event One Day Mean -0.000392304 -0.00028924Known Variance 0.000146349 9.89339E-05Observations 195 906Hypothesized Mean Difference 0z -0.111155694P(Z<=z) one-tail 0.455746442z Critical one-tail 1.644853627P(Z<=z) two-tail 0.911492885z Critical two-tail 1.959963985

z-Test: Two Sample for Means

Negative CRR 2 Days Negative No Event CRR 2 DaysMean 0.000478319 -1.13612E-05Known Variance 7.64484E-05 5.03663E-05Observations 195 906Hypothesized Mean Difference 0z 0.731897176P(Z<=z) one-tail 0.232115665z Critical one-tail 1.644853627P(Z<=z) two-tail 0.46423133z Critical two-tail 1.959963985

35

Table 16

As it can be seen above in the given table, the results from the Z test for a given alpha

of 10% do not vary from those already mentioned above with a given alpha of 5%. (Table 16)


Negative One Day Negative No Event One Day Mean -0.000392304 -0.00028924Known Variance 0.000146349 9.893E-05Observations 195 906Hypothesized Mean Difference 0z -0.111155694P(Z<=z) one-tail 0.455746442z Critical one-tail 1.281551566P(Z<=z) two-tail 0.911492885z Critical two-tail 1.644853627


Negative CRR 2 Days Negative No Event CRR 2 DaysMean 0.000478319 -1.13612E-05Known Variance 7.64484E-05 5.03663E-05Observations 195 906Hypothesized Mean Difference 0z 0.731897176P(Z<=z) one-tail 0.232115665z Critical one-tail 1.281551566P(Z<=z) two-tail 0.46423133z Critical two-tail 1.644853627

36

4.2.1.2. Positive Events Table 17

In the case for the Positive Events’ Z test analysis for a given alpha of 5%, the results

above demonstrate that the difference between the mean of the Positive One Day and the

Positive No Event One Day is not indistinguishable from 0. The already mentioned is

confirmed in the P-Value for one tail and the obtained z. (Table 17)

As well as, in the case for the Positive CRR 2 Days in comparison with the Positive No

Event CRR 2 Days, the result is the same: the difference in means is not indistinguishable from

0.


Positive One Day Positive No Event One DayMean 0.002153877 0.000710625Known Variance 0.000165333 8.17966E-05Observations 201 1018Hypothesized Mean Difference 0z 1.518873825P(Z<=z) one-tail 0.064397129z Critical one-tail 1.644853627z Critical two-tail 1.959963985


Positive CRR 2 Days Positive No Event CRR 2 Days Mean 0.001518644 0.000426227Known Variance 9.42821E-05 4.45093E-05Observations 201 1018Hypothesized Mean Difference 0z 1.525526479P(Z<=z) one-tail 0.063563917z Critical one-tail 1.644853627P(Z<=z) two-tail 0.127127834z Critical two-tail 1.959963985

37

Table 18

In the case for a given alpha of 10%, the results differ from those with an alpha of 5%.

In this case, the mean in both cases, the Positive One Day and the Positive CRR 2 Days are

statistically larger than the means of the Positive No Event. (Table 18)

4.2.2. Grouping 2 In the case for Grouping 2, the results are the same for the given statistical Z tests as in

Grouping 1. However, even if results are the same in term of the Z tests, they are not for the

rest of the statistical information provided for reference; such as means, variances, and z’s.

Therefore, results are shown below with the purpose of providing the reader the rest of the

statistical data for any further reference and/or interpretation.

Table 19


Positive One Day Positive No Event One DayMean 0.002153877 0.000710625Known Variance 0.000165333 8.17966E-05Observations 201 1018Hypothesized Mean Difference 0z 1.518873825P(Z<=z) one-tail 0.064397129z Critical one-tail 1.281551566z Critical two-tail 1.644853627



Number of Events Total= 2321 124 5.343% 978 42.137% 1097 47.264% 122 5.256%



Prior Day Returns NEG<=-1.5% -1.5%<NNE<0% 0%<PNE<1.5% 1.5%<=PE

38



Negative One Day Negative No Event One Day Mean -0.00013404 -0.000329508Known Variance 0.000154937 0.000101298Observations 124 977Hypothesized Mean Difference 0z 0.168034889P(Z<=z) one-tail 0.433277916z Critical one-tail 1.644853627P(Z<=z) two-tail 0.866555832z Critical two-tail 1.959963985


Negative CRR 2 Days Negative No Event CRR 2 DaysMean 0.000498963 2.16043E-05Known Variance 8.67615E-05 5.09776E-05Observations 124 977Hypothesized Mean Difference 0z 0.550522004P(Z<=z) one-tail 0.290980694z Critical one-tail 1.644853627P(Z<=z) two-tail 0.581961389z Critical two-tail 1.959963985

39

Table 21


Negative One Day Negative No Event One Day Mean -0.00013404 -0.000329508Known Variance 0.000154937 0.000101298Observations 124 977Hypothesized Mean Difference 0z 0.168034889P(Z<=z) one-tail 0.433277916z Critical one-tail 1.281551566P(Z<=z) two-tail 0.866555832z Critical two-tail 1.644853627


Negative CRR 2 Days Negative No Event CRR 2 DaysMean 0.000498963 2.16043E-05Known Variance 8.67615E-05 5.09776E-05Observations 124 977Hypothesized Mean Difference 0z 0.550522004P(Z<=z) one-tail 0.290980694z Critical one-tail 1.281551566P(Z<=z) two-tail 0.581961389z Critical two-tail 1.644853627

40



Positive One Day Positive No Event One DayMean 0.002397801 0.000787433Known Variance 0.000159162 8.85584E-05Observations 122 1097Hypothesized Mean Difference 0z 1.368194806P(Z<=z) one-tail 0.085625554z Critical one-tail 1.644853627P(Z<=z) two-tail 0.171251107z Critical two-tail 1.959963985



41

Table 23

4.3. Magnitude of the Positive and Negative Market Reactions Post Event Finally, a third and last analysis’ results are provided in this section with the purpose of

assessing the market reactions post Negative and Positive Events for a different Grouping of

data (Grouping 3), which is already explained in the data and methodology chapter.

In the following table (24) are the results from the post event returns already split into

positive and negative reactions, meaning the positive those returns above 0, and negative those

below 0. As well as the number of days with a positive reaction and those with a negative

reaction post event. Also, the means of the already mentioned returns from one day after the

events divided into positive and negative returns for both Positive and Negative Events.


Positive One Day Positive No Event One DayMean 0.002397801 0.000787433Known Variance 0.000159162 8.85584E-05Observations 122 1097Hypothesized Mean Difference 0z 1.368194806P(Z<=z) one-tail 0.085625554z Critical one-tail 1.281551566P(Z<=z) two-tail 0.171251107z Critical two-tail 1.644853627


Positive CRR 2 Days Positive No Event CRR 2 Days Mean 0.001727047 0.00048172Known Variance 0.000086213 4.90152E-05Observations 122 1097Hypothesized Mean Difference 0z 1.436691913P(Z<=z) one-tail 0.075402777z Critical one-tail 1.281551566P(Z<=z) two-tail 0.150805554z Critical two-tail 1.644853627

42

Table 24

In the following two sections graphics are provided with the purpose of giving the

reader a visual aid to graphically analyse the positive reaction post event and the negative

reaction post event for each event classification. In each section comes first the table for the

given event, in which the means of the returns from the days of the events are provided. As

well as, the two means of the events’ following days’ returns (Post Event Positive and Negative

Reactions). In the same table comes also de percentage in which the post event reactions are

distributed. Which can be seen as the chance of occurrence of either a negative post event

reaction or a positive post event reaction.

Furthermore, tables which provide information on the magnitude of the reactions are

displayed in order to better understand the graphics. Each table of the “Magnitude of

Reactions” is classified in either Positive or Negative, depending on the case of analysis.

Important to realize, in the tables already mentioned above and for the graphics, the first input

is the mean from the day of the event, henceforth, the second (Post Event in the table) is the

mean of the event plus the mean of the post event reactions (either positive or negative,

depending on the case). This can help to analyse the post event reactions while taking into

consideration also the magnitude of the event itself.


Groping 3

% Mean n % Mean nPositive,Reactions %RR1>0 51.282% 0.8549% 100 55.224% 1.0358% 111Negative,Reactions %RR1<0 48.718% =0.9804% 95 44.776% =0.7964% 90


Number of Events

Market,Reactions,Mean of the Event !1.9391% 1.9155%

195 201Prior Day Returns NEG<=-1.2% 1.2%<=PE

Chance'of'Ocurrence't= Post'One'Day'RR'Means Event N %'of'Ocurrence'

0 :0.019391044 Negative'Event 195 11 0.008548872 Post'Event'Positive'Reactions 100 0.5128205131 :0.009804068 Post'Event'Negative'Reactions 95 0.487179487

Negative(Event(

43

Table 25 provides the data for the Negative Event’s analysis post event reaction. As

can be seen the means of the returns from the days of the events are provided. As well as, the

two means of the events’ following days’ returns (Post Event Positive and Negative Reactions).

In the same table comes also de percentage in which the post event reactions are distributed.

Table 26

Figure 1

Figure 1 and Table 26 display the observed positive post Event market reaction for the

negative price shocks, in which the mean reversion to the mean pattern can be assessed.

0 0.0000%Event 1 *1.9391%

Post0Event0 2 *1.0842%

Magnitude0of0Positive0Reactions0

-1.9391%

-1.0842%

0.0352122% 0.0352122% 0.0352122%

-2.5000%

-2.0000%

-1.5000%

-1.0000%

-0.5000%

0.0000%

0.5000%

0 1 2

Mea

n of

Ret

urns

Event and Post Event

Post Negative Event Positive Market ReactionsMagnitude of Positive Reactions Total Mean for Graphics use

44

Table 27

Figure 2

Figure 2 and table 27 provide information of the positive post Event market reaction in

the case for negative price shocks, in which the price drift in the same direction of the price

shock can be seen.


Table 28 provides the data for the Positive Event’s analysis post event reaction. As can

be seen the means of the returns from the days of the events are provided. As well as, the two

0.0000% 0 0.0000%$1.9391% Event 1 $1.9391%$1.0842% Post3Event3 2 $2.9195%

Magnitude3of3Positive3Reactions3 Magnitude3of3Negative3Reactions3

-1.9391%

-2.9195%

0.0352122% 0.0352122% 0.0352122%

-3.5000%

-3.0000%

-2.5000%

-2.0000%

-1.5000%

-1.0000%

-0.5000%

0.0000%

0.5000%

0 1 2

Mea

n of

Ret

urns


Post Negative Event Negative Market Reactions

Magnitude of Negative Reactions Total Mean for Graphics use

Chance'of'Ocurrence't= Post'One'Day'RR'Means Event N %'of'Ocurrence'0 0.01915507 Positive'Event 201 11 0.010357579 Post'Event'Positive'Reactions 111 0.5522388061 D0.007964022 Post'Event'Negative'Reactions 90 0.447761194

Positive(Event(

45

means of the events’ following days’ returns (Post Event Positive and Negative Reactions). In

the same table comes also de percentage in which the post event reactions are distributed.

Table 29

Figure 3

Figure 3 and table 29 shown above give the results of the positive post event market

reaction in the case of positive price shocks. As can be seen in figure 3, there is a price drift in

the same direction of the price shock.

0 0.0000%Event 1 1.9155%

Post/Event/ 2 2.9513%

Magnitude/of/Positive/Reactions/

1.9155%

2.9513%

0.0352122%

0.0352122% 0.0352122%0.0000%

0.5000%

1.0000%

1.5000%

2.0000%

2.5000%

3.0000%

3.5000%

0 1 2

Mea

n of

Ret

urns


Post Positive Event Positive Market Reactions

Magnitude of Positive Reactions Total Mean for Graphics use

46

Table 30

Figure 4

As provided above in figure 4 and table 30, it can be said that there is a mean reversion

to the mean pattern for the negative post event market reactions in the case of the positive price

shocks.

0.0000% 0 0.0000%1.9155% Event 1 1.9155%2.9513% Post1Event1 2 1.1191%

Magnitude1of1Negative1Reactions1Magnitude1of1Positive1Reactions1

1.9155%

1.1191%

0.0352122% 0.0352122% 0.0352122%0.0000%

0.5000%

1.0000%

1.5000%

2.0000%

2.5000%

0 1 2

Mea

n of

Ret

urns


Post Positive Event Negative Market Reactions

Magnitude of Negative Reactions Total Mean for Graphics use

47

5. Results Analysis In this section, the analysis and comments on the already given results in the previous chapter

will be provided. This analysis is in order to gauge the market reactions post price shocks from

the three different studies done in the thesis.

5.1.1. Single Mean Hypothesis Testing, Z test: One Sample Mean As it is already mentioned above, results from the Z test do no vary between the post event one

day and the post event two days. As well as results do not vary between Grouping 1 and

Grouping 2, either from one day and two days’ analysis. In both cases the means from the

negative events and negative no events are not indistinguishable from cero. In like manner, the

results are the same for the two given alphas, 0.05 and 0.1 used in this analysis.

Taking into consideration the given results, this statistical test would suggest that there

is no supporting evidence of significant market reactions following either negative events nor

negative no events. Hence, this statistical test suggests that there is a very poor market reaction

following a negative price shock

On the other side, for both cases, one and two days’ analysis for the positive events and

the positive no events the results show that the means are indistinguishable from cero; more

over, the means are statistically higher than cero in all cases. This result would suggest there is

a positive market reaction following positive events. However, this is not enough evidence for

suggesting either under reaction nor momentum following a positive price shock.

5.1.2. Two Sample Mean Hypothesis Testing, Z test: Two Sample Means First, for the analysis of the negative events, the results demonstrate that the difference between

the mean of the Negative One Day and the Negative No Event One Day is not indistinguishable

from 0. As well as, in the case for the Negative CRR 2 Days in comparison with the Negative

No Event CRR 2 Days, the result is the same: the difference in means is not indistinguishable

from 0. In the same fashion, the results from the Z test for a given alpha of 10% do not vary

from those with an alpha of 5%.

Therefore, with attention to the already seen results from this statistical test, in the case

of the Negative Events, there is no significant evidence suggesting either overreaction nor

under reaction following a negative price shock in the Mexican Capital Market for the given

period of time used in this analysis. On the contrary, when assessing just the mean on its own,

48

the evidence suggests a poor market reaction following a negative price shock. Finally, this

statistical test would suggest there is no difference in terms of the magnitude of market

reactions following either a negative price shock nor a negative no event price shock.

On the other side, in the case for the Positive Events’ Z test analysis for a given alpha

of 5%, the results above demonstrate that the difference between the mean of the Positive One

Day and the Positive No Event One Day is not indistinguishable from 0. As well as, the results

are the same for the Positive CRR 2 Days in comparison with the Positive No Event CRR 2

Days.

Different from the Negative Events analysis, in the statistical test with a given alpha of

10%, the results differ from those with an alpha of 5%. Taking into account an alpha of 10%,

the means in both cases, the Positive One Day and the Positive CRR 2 Days are statistically

larger than the means from the Positive No Event.

Given the obtained results, it can be said that in the case for the Positive Events, there

is no significant evidence supporting stronger positive market reactions following a positive

price shock than when there is a positive no event price shock for a given alpha of 5%.

However, when taking the statistical with an alpha of 10%, the statistical tests confirm that the

means for the Positive Event’s case are larger than from those of the Positive No Events. This

would suggest that the market reactions are stronger in the same directions of the event

following a positive price shock than when there is no significant positive price shock. Hence,

from a statistical point of view, the null hypothesis can be rejected in the case of positive events,

and thereafter consider the suggested behavioral finance causes from the hypothesis testing

table already displayed earlier. Nonetheless, when taking into consideration just the mean on

its own, the evidence suggests a poor reaction following a positive price shock. Since the

positive price shock range is at least equals or above 1.2%, and the mean of the following days

of the price shocks is very low (0.21%). This would indicate a poor reaction, which implies

that there is not enough evidence in terms of magnitude when comparing the mean against the

Positive Event Range of 1.2%. Finally, the behavioral finance suggested causes on the

hypothesis testing can not be accepted taking into consideration the already mentioned in terms

of magnitude. Thus, it can be said there is just a mild positive market reaction following a

positive price shock.

Note: In the case of Grouping 2, the results are the same for the given statistical Z tests as in

Grouping 1. Thus, there is no need of providing result analysis again for Grouping 2 since the

analysis would be the same as in for Grouping 1.

49

5.1.3. Magnitude of the Positive and Negative Market Reactions Post Event In the case of this third analysis, with the purpose of assessing the market reactions post

Negative and Positive Events for a different Grouping of data (Grouping 3), the results

completely differ from the two statistical tests already mentioned above in terms of providing

supporting evidence for Over and Under reaction in the Mexican capital market.

When analysing the Negative Events, its been found that the post event market reactions

can not be attributed to only one type of reaction (over or under reaction) since the reactions

following a negative price shock can be either positive or negative, depending on each case.

For the given period of analysis, its been found that about a bit more than half of the times the

market reaction following a negative price shock is positive; likewise, the other half of the

times the market reaction is negative. To put it into numbers, 51.2% of the times the reaction

is positive, while 48.8% of the times the reaction is negative.

In the case for the positive reactions, when analysing the magnitude of the mean of the

negative events itself (-1.94%) in comparison to the mean of the positive reactions (0.86%),

the positive reaction is found to be very significant, it represents about half of the mean from

the negative events in terms of magnitude. This result would suggest prove of overreaction for

negative events as following a negative price shock there is a reversal to the mean pattern for

the given cases of positive market reactions.

On the other hand, in the case for the negative reactions its been found that those are

stronger than the positive reactions following a negative price shock when anchoring the

comparison to the mean of the negative event (-1.94%). The mean of the negative reactions (-

0.98%) is more than half of the mean of the negative events. Since prices would continue to

fall following a negative price shock, this would suggest evidence of momentum and under

reaction following a negative price shock for the given cases of negative market reactions.

Note: For graphic reference of the mentioned in the two paragraphs above see 4.3.1.1.

At the same time, when analysing the Positive Events, its been found that the post event

market reactions can not be attributed to only one type of reaction (over or under reaction)

since the reactions following a negative price shock can be either positive or negative,

depending on each case. For the given period of analysis, its been found that slightly more than

half of the times the market reaction following a positive price shock is positive; likewise,

slightly less than half of the times the market reaction is negative. To put it into numbers, 55.2%

of the times the reaction is positive, while 44.8% of the times the reaction is negative.

50

When analysing the positive reactions, the magnitude of the mean of the positive events

itself (1.92%) in comparison to the mean of the positive reactions (1.04%), the positive reaction

is found to be very significant, it represents more than half of the mean of the positive events

in terms of magnitude. This result would suggest prove of momentum and under reaction

following a positive price shock since there is a positive price drift towards the same direction

of the event for the given cases of positive market reactions.

On the other hand, in the case for the negative reactions, its been found that those are

weaker than the positive reactions following a positive price shock when anchoring the

comparison to the mean of the positive event (1.92%). The mean of the negative reactions (-

0.80%) is less than half of the mean of the positive events. Since prices would drop following

a positive price shock, this would suggest evidence of overreaction for positive events as

following a positive price shock there is a reversal to the mean pattern for the given cases of

negative market reactions.

Note: For graphic reference of the mentioned in the two paragraphs above see 4.3.1.2.

51

“Investors should remember that excitement and expenses are their enemies. And if they

insist on trying to time their participation in equities, they should try to be fearful when

others are greedy and greedy when others are fearful.”

“Cash combined with courage in a time of crisis is priceless.”

Warren Buffett

6. Conclusions

“The fact that people will be full of greed, fear or folly is predictable. The sequence is not

predictable.”

Warren Buffett

After the deep research, analysis, and long learning process on which this thesis is based, there

is a need to retrieve what is been said in the beginning: In the field of investments, the

implications of behavioural finance are substantial, analysing investor behaviour is an essential

key in order to understand the sometimes rare and unexplainable fluctuations of financial

markets. Hence, understanding behavioural finance can provide significant advantages for

investors to the end of taking better investment decisions. Optimal decision making for

investing in the stock market requires a deep understanding of investors rationality and more

important irrationality in a global perspective. Behavioral Finance is an integral part of the

decision-making process because it heavily influences the investors’ performance.

Generally speaking, it can be said that markets are not fully efficient, moreover, several

studies have fail to the same conclusion, market participants are not fully rational. The idea of

having market participants whom account for the characteristics of full rationality are rarely

seen. Certainly, new information is not instantaneously and fully incorporated into stock prices

in an unbiased fashion. Thus, prices are likely not to be fair in al cases.

The most compelling evidence suggests that market participants are not fully rational.

Actually, there have been considerable studies which advocate stock crashes and booms to

human errors, proposing those errors are influenced by behavioral finance bias. After the

research gone through, it can be said that market participants have a strong tendency to fail into

the psychological bias already explained in this thesis, studies suggest there is a Systematic

error in judgment and decision-making common to all human beings which can be due to

cognitive limitations, motivational factors, and/or adaptations to natural environments. The

mention cognitive bias have their roots in Prospect Theory and Heuristics theory. The

52

conducted studies on these two theories support the following human cognitive bias:

Anchoring, Loss Aversion, Frame Dependence, Mental Accounting, Misperceiving

Randomness, Herding, and Overconfidence.

Studies propose the mentioned behavioral theories and bias to be the causes of

overreaction and under reaction in capital markets. In the same manner, the roots of contrarian

investing and momentum investing. As a result of this, prices may well not represent their fair

value at times due to these market reactions in the capital market. As well as, its been found

that in some cases there is momentum in the short run but mean reversion in the long term.

Stock prices tend to first drift (sign of under reaction), but thereafter prices correct (sign of

overreaction).

In the case of overreaction and contrarian investing there are tons of research providing

evidence proposing that often times winner stocks are over bought (Above fair value) and loser

stocks are oversold (under fair value), hence profits occur as winner stocks retreat to their fair

value and loser stocks rally to their fair value. As well as, empirical studies have also found

examples of under reaction. Causing momentum in prices as stock prices keep on drifting in

the same direction as the information is slowly incorporated into prices. From this view, prior

winners outperform and prior losers underperform. The evidence proposes that under reaction

and price drifts are usually related to important public announcements. Therefore, positive

news lead to subsequent positive abnormal returns, and negative news lead to subsequent

negative abnormal returns.

Although the mentioned above may be true, one possibility to explain these market

anomalies are the deviations to be expected under the Efficient Market Hypothesis. However,

this thesis aimed to provide information suggesting that these market anomalies are not caused

by those deviations under market efficiency. A central theme on this work was to prove that

prices sometimes overreact or underreact to some signal. Thus, the main questioning is What

is the role of Behavioral Finance in the capital market, and what are its implications for price

shocks?

First, when taking into consideration the evidence on Prospect theory, it could be said

that if investors behave as explained in prospect theory, they would react differently to positive

and negative price shocks. This theory suggests that following significant gains, such as a

strong positive price shocks, investors would become more risk averse, thereafter, they would

sell in order to obtain profits. On the contrary, this theory implies that investors tend to become

more risk seeking following losses. Thus, instead of selling after a negative price shock,

53

investors would hold the stock or even buy more with the purpose of gambling back their

losses.

Second, when studying overreaction and contrarian investing, it would imply the initial

price shock to be too strong mainly caused by emotions on the day of a price shock. As a result,

prices would drift towards unfair value on the event day, hence, prices should fall following

positive events and raise after negative events in order to return to fair value.

Third, when taking into consideration the evidence in reference to under reaction and

momentum, it would imply a price drift following a price shock in the same direction of the

price shock. The magnitude and length of the period of time for the price drift would be mainly

attributed to the magnitude of the new information’s impact which causes the original price

shock. At the same time, this would cause momentum following price shocks.

Finally, a complete explanation of markets’ post reaction to price shocks would be

causal attribution. This explains the market reaction in relationship to whether the investors

attribute the price shock to a stable or non- stable cause. If investors attribute the cause of the

event is stable, they would believe the past results will repeat. As a consequence, positive

events would lead investors to expect more gains and increase optimism, while on the other

side, negative events would lead investors to expect more losses and increase pessimism. On

the other hand, when investors believe the cause of the event is non-stable, it would imply

positive market reactions following negative events, and negative reactions following positive

events.

Thereafter, this thesis displays three different analyses in order to suggest proof the

main hypothesis of this thesis: Markets tend to overreact to bad news, and underreact to good

news. The three analyses provided were done on the Mexican Capital Market main index (IPC)

with a time frame from 2009 to 2018. First, a single mean hypothesis testing, with the purpose

to asses whether the post event reaction was positive or negative. Second, two sample mean

hypothesis testing, which had the target of gauging whether there was supporting evidence on

Overreaction for Negative Events and Under reaction for positive events when comparing the

means of the Events in contrast to the No Events. Finally, a third analysis based on the means

of the price shocks and the post one-day reaction in order to assess the magnitude of the positive

and negative market reactions post event. The function of this last analysis was to asses the

market reactions post event, focusing merely on post price shocks reactions, both positive and

negative, thereafter, the post event market reactions can be judged in terms of their relationship

to behavioral finance theories.

54

First, the single mean hypothesis testing suggests that following a negative price shock

the statistical evidence suggests a poor market reaction, thus, this evidence would propose that

the market incorporates the bad news on the same day of the event. On the other hand, in the

case for the positive price shocks, the evidence advocates a mild positive market reaction

following positive events, which is not enough evidence to propose signs of either under

reaction nor momentum following positive price shocks.

Similarly, the two sample mean hypothesis testing failed to the same conclusions in the

analysis of the negative events, there is no significant evidence on either overreaction nor under

reaction in the case of negative price shocks. The evidence suggests a poor market reaction

following a negative price shock. Finally, this test would again suggest that there is no

difference in terms of the magnitude of market reactions following either a negative price shock

nor a negative no event price shock.

On the side, in the case of positive events, in the statistical test with a given alpha of

10%, the market reactions are stronger in the same directions of the event following a positive

price shock than when there is no significant positive price shock. Thereafter, from a statistical

point of view, the null hypothesis applied to this test could be rejected in the case of the positive

events. Hence, the suggested behavioral finance causes from the hypothesis testing table

already displayed earlier on this thesis could be accepted. Nonetheless, when taking into

consideration just the mean on its own, the evidence suggests a poor reaction following a

positive price shock. Finally, this test does not provide enough evidence on under reaction

following a positive price shock, and it can be said there is just a mild positive market reaction

following a positive price shock.

Although it may be true that sometimes there are just mild market reactions following

strong price shocks, this thesis aims to prove the contrary. The assumption that markets react

poorly following strong price shocks truly contradicts all the provided evidence in the literature

review.

In contrast to the mentioned tests, the results from the third analysis completely differs

in terms of providing supporting evidence for Over and Under reaction. When analysing the

market reactions, it was found that in both cases, the Negative Events and Positive Events, the

event market reactions can not be attributed to only one type of reaction (over or under reaction)

since the reactions following a significant price shock can be either positive or negative. It was

found that about slightly more than half of the times the market reactions following a price

shock are positive; likewise, the other half of the times the market reactions are negative.

55

When analysing the negative price shocks from the days of the events, the two market

post event reactions, both positive and negative account to be very significant. The positive

reactions mean represents about half of the mean from the negative events. This result would

suggest proof of overreaction for negative events following a negative price shock since there

is a reversal to the mean pattern for the given cases of positive market reactions. On the other

hand, in the case for the negative reactions its been found that those are stronger than the

positive reactions following a negative price shock when anchoring the comparison to the mean

of the negative event. The mean of the negative reactions is more than half of the mean of the

negative events. Since prices would continue to fall following a negative price shock, this

would also suggest evidence of momentum and under reaction following a negative price shock

for the given cases of negative market reactions.

On the side, for the positive price shocks, the two market post event reactions, both

positive and negative post event account to be very significant. The positive reactions account

for a mean of more than half of the mean of the positive events. This result would suggest proof

of under reaction and momentum following a positive price shock since there is a positive price

drift towards the same direction of the event for the given cases of positive market reactions.

On the other hand, in the case for the negative reactions, its been found that those are weaker

than the positive reactions following a positive price shock when anchoring the comparison to

the mean of the positive event. The mean of the negative reactions is less than half of the mean

of the positive events. Since prices would drop following a positive price shock, this would

suggest evidence of overreaction for positive events since following a positive price shock

there is a reversal to the mean pattern for the given cases of negative market reactions.

Finally, in the light of this last analysis, the evidence suggests proof of both under and

overreaction in the Mexican stock capital market. On the whole, the market tends to both

overreact and underreact to price shocks, in both cases, positive and negative price shocks. In

essence, when there is a negative price shock, the market tends to overreact a bit more than half

of the times, while the rest of the times underreacts. In the same order, it suggests contrarian

investing and momentum. While on the other side, when there is a positive price shock, the

market underreacts slightly more than half of the times while the rest of the times overreacts.

Likewise, in the same manner it suggests momentum slightly more than half of the times while

the rest of the times it can suggest contrarian investing.

Finally, all in all, to put a very simple explanation of the market possible reaction to

price shocks comes the following process interpretation. When there is new substantial

information, the market may either underreact or overreact to this new information causing a

56

price shock. The price shock may be either negative or positive, depending on the new

information. Greed comes into play when good new information arrives while panic and fear

do so when bad news arrive. This price shock will lead to substantial post event reactions which

can be either positive or negative. In the case of negative price shocks, if the market

underreacted the post event reaction will be negative; on the contrary, if the market overreacted,

the post event reaction will be positive. On the other side, in the case of a positive price shock,

if the market underreacted the post event reaction will be positive; on the other hand, if the

marked overreacted, the post event reaction will be negative.

Finally, there are various different approaches to analyse the capital market from a

behavioral finance perspective in contrast to the efficient market hypothesis. After deep

research and learning process gone through this thesis, a few different possible topics for

further research emerged. Among the ideas for further research, it would be very interesting to

build up portfolios from the Mexican capital market based on momentum and/or contrarian

investment strategies with the purpose of evaluating its performance against the market. As

well as, do the same analyses from this thesis but in single securities and in different groups of

securities.

57

Reference List Alejandra Arteaga (2015). Los días negros en la historia de las bolsas. Negocios Milenio.com.

Available in http://www.milenio.com/negocios/dias_negros_historia_bolsas_valores-

las_bolsas_de_valores-caida_bursatil_0_578942328.html

Andersson, M., Hedesstrom, M., & Garling, T. (2014). A social-psychological perspective on

herding in stock markets. Journal of Behavioral Finance, 15(3), 226-234.

doi:10.1080/15427560.2014.941062

Andreassen, P. & Kraus, S. (1988). Judgmental prediction by extrapolation. Department of

Psychology, Harvard University

Atkins, A. B., & Dyl, E. A. (1990). Price Reversals, Bid-Ask Spreads, and Market

Efficiency. Journal of Financial and Quantitative Analysis, 25(4), 535-547.

DOI: 10.2307/2331015

Banerjee, Arindam (September, 2011). Application of Behavioural Finance in Investment

Decisions: An Overview The Management Accountant, Vol. 46, No. 10, October 2011.

Available at SSRN: https://ssrn.com/abstract=2118972

Barber, Brad M., and Odean, Terrance (1999). The Courage of Misguided Convictions.

Financial Analysts Journal 55(6).

Barberis, Nicholas, Andrei Shleifer, And Robert Vishny (1998). A moderl of investor

sentiment. Journal of Financial Economics 49, 307-343.

Bernard, Victor L., Thomas, Jacob K. (1990)."Evidence that stock prices do not fully reflect

the implications of current earnings for future earnings." Journal of Accounting and

Economics 13(4): 305-340. Available in http://hdl.handle.net/2027.42/28288

Bremer, M. and Sweeney, R.J., (1991). The reversal of large stock-price increases. Journal of

Finance 46, pp. 747-754.

Brown, Lawrence D. (1997). Earnings surprise research: Sysnthesis and perspectives. Financial

Analysts Journal 53, 13-19.

58

Chen, Andrew H. Thomas F. Siems (2002). “An Empirical Analysis of the Capital Markets’

Response to Cataclysmic Events.” Southern Methodist University, School of Business,

Working Paper, Dallas.

Choi, J. J., Laibson, D., & Madrian, B. C. (2009). Mental accounting in portfolio choice:

Evidence from a flypaper effect. The American Economic Review, 99(5), 2085- 2095.

doi:10.1257/aer.99.5.2085

Cox and Pertson (1994). Stock returns following large one-day declines: evidence on short-

term reversals and longer-term performance, Journal of Finance 49 (1994), pp. 255-

267.

David Blake (2016). Investopedia. The Top 17 Investing Quotes of All Time. Retrived from

https://www.investopedia.com/financial-edge/0511/the-top-17-investing-quotes-of-

all-time.aspx

De Bondt, W.F. and R.H. Thaler (1985) “Does the Stock Market Overreact?” Journal of

Finance, 40: 793-805.

Doviak, P. (2016). Applying behavioral finance to ourselves. Journal of Financial Planning,

29(6), 46. Retrieved from https://www.onefpa.org/journal/Pages/default.aspx

Duxbury, D. (2015). Behavioral finance: Insights from experiments II: Biases, moods and

emotions. Review of Behavioral Finance, 7(2), 151-175. doi:10.1108/RBF-09- 2015-

0037

Fama, Eugene F. (1965). Random Walks in Stock Market Prices. Financial Analysts Journal

21 (5): 55 - 59.

Fama, Eugene F., 1998, Market efficiency, long-term returns and behavioral finance, Journal

of Financial Economics 49.

Greenwood, R., & Nagel, S. (2009). Inexperienced investors and bubbles. Journal of Financial

Economics, 93(2), 239-258. doi:10.1016/j.jfineco.2008.08.004

Grinblatt, Mark, Ronald W. Masulis, and Sheridan Titman (1984). The valuation effects of

stock splits and stock dividends. Journal of Financial Economics 13, 97-112.

59

Heath, C., & Tversky, A. (1991). Preference and belief: Ambiguity and competence in choice

under uncertainty. Journal of Risk and Uncertainty, 4(1), 5-28.

doi:10.1007/BF00057884

Hirshleifer, D. (2015). Behavioral finance. Annual Review of Financial Economics, 7(1), 133-

159. doi:10.1146/annurev-financial-092214-043752

Hong, Harrison, and Jeremy C Stein. (1999). “A Unified Theory of Underreaction, Momentum

Trading and Overreaction in Asset Markets.” Journal of Finance LIV (6): 2143-2184.

Ikenberry. David, Josef Lakonishok, and Theo Vermaelen (1995). Market Underreaction to

open market share repurchases. Journal of Financial Economics 39, 181-208.

Jegadeesh, N. and S. Titman (1993). Returns to Buying Winners and Selling Losers:

Implications for Stock Market Efficiency. Journal of Finance, 48 (1): 65-91.

Jordan, B., Miller, T., & Dolvin, S. (2015). Fundamentals of investments: Valuation and

management (7th ed.). New York, NY: McGraw-Hill Education.

Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk.

Econometrica, 47(2), 263-291. doi:10.2307/1914185

Kahneman, D., Slovic, P., & Tversky, A. (Eds.). (1982). Judgment under Uncertainty:

Heuristics and Biases. Cambridge: Cambridge University Press.

doi:10.1017/CBO9780511809477

Kahneman, Daniel. and Tversky, Amos. 1979. Prospect Theory: An Analysis of Decision under

Risk. Econometrica 47(2): 263 – 291.

Kent Daniel, David Hirshleifer, and Avanidhar Subrahmanyam (1998). Investor Psychology

and Security Market Under-‐ and Overreactions. The Journal of Finance.

https://doi.org/10.1111/0022-1082.00077

Kim, I. (2002). An analysis of the market reward and torpedo effect to firms that consistently

meet expectations. Working paper. University of Michigan.

Kliger, D., van den Assem, M., & Zwinkels, R. (2014). Empirical behavioral finance. Journal

of Economic Behavior & Organization, 107, 421-427. doi:10.1016/j.jebo.2014.10.012

60

Larson, S and J. Madura, 2002. What drives stock price behavior following extreme one-day

returns? Journal of Financial research.

Lisa Smith (2017). Investopedia, Modern Portfolio Theory Vs. Behavioral Finance. Available

in https://www.investopedia.com/articles/investing/041213/modern-portfolio-theory-

vs-behavioral-finance.asp

Lowe, Logan S. (2008). Short Term Responses of the Stock Market to Large Daily Price

Changes: An Empirical, Behavioral Finance, and Neuroeconomic Analysis. Princeton

University Senior Theses. URI: http://arks.princeton.edu/ark:/88435/dsp015999n4304

Michael Johnston (2014). Dividend.com. 41 Inspiring And Intelligent Investing Quotes.

Retrived from http://www.dividend.com/dividend-education/41-inspiring-and-

intelligent-investing-quotes/

Michael Kaestner (2006). Investors’ Misreaction to Unexpected Earnings: Evidence of

Simultaneous Overreaction and Underreaction. Working paper. Montpellier University.

http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.175.7118&rep=rep1&type=

pdf

Nofsinger, John R. (2001). Investment madness: how psychology affects your investing – and

what to do about it. USA: Pearson Education

Parikh (2011). Value Investing and Behavioral Finance. New Delhi : Tata Mcgraw Hill.

Poterba, James M., and Lawrence Summers (1998). Mean reversion in stock returns: Evidence

and Implications. Journal of Financial Economics 22, 27-59.

Power Finance Quotes (2018). Behavioral Fiance. Available in

https://powerfinancequotes.wordpress.com/add-or-update-your-own-

quote/behavioral-finance/

Roider, A., & Voskort, A. (2016). Reputational herding in financial markets: A laboratory

experiment. Journal of Behavioral Finance, 17(3), 244-266.

doi:10.1080/15427560.2016.1203322

S&P/ Dow Jones Indices (2018). S&P/BMV IPC. Data Series. Retrieved from

https://latam.spindices.com/indices/equity/sp-bmv-ipc

61

Sewell, Martin. (2007). Behavioral Finance. University of Cambridge

http://www.behaviouralfinance.net/behavioural-finance.pdf.

Shiller, R. J. (2003). From efficient markets theory to behavioral finance. The Journal of

Economic Perspectives, 17(1), 83-104. doi:10.1257/089533003321164967

Sinha, P. C. (2015). Stocks’ pricing dynamics and behavioral finance: A review. Management

Science Letters, 5(9), 797-820. doi:10.5267/j.msl.2015.7.003

Statman, M. (2014). Behavioral finance: Finance with normal people. Borsa Istanbul. Review,

14(2), 65-73. doi:10.1016/j.bir.2014.03.001

Sturm, Ray R. (2003). “Investor Confidence and Returns Following Large One-Day Price

Changes.” Journal of Finance. 201-216.

Thaler, R. (1980). Toward a positive theory of consumer choice. Journal of Economic Behavior

& Organization, 1(1), 39-60. doi:10.1016/0167-2681(80)90051-7

Thaler, R. H. (1999). The end of behavioral finance. Financial Analysts Journal, 55(6), 12-17.

doi:10.2469/faj.v55.n6.2310

Thaler, R. H. (Ed. Volume II). (1993). A SURVEY OF BEHAVIORAL FINANCE. Nicholas

Barberis and Richard Thaler. Advances in behavioral finance. New York, NY, US:

Russell Sage Foundation. Chapter 1, p 1.

Thaler, R. H. (Ed. Volume II). (1993). Advances in behavioral finance. New York, NY, US:

Russell Sage Foundation. Preface: xi.

Thaler, R., & Barberis, N. (2002). A survey of behavioral finance. Handbook of the Economics

of Finance, National Bureau of Economic Research, 1051-1121. Retrieved from

http://faculty.som.yale.edu/nicholasbarberis/ch18_6.pdf

Thornton McEnery (2011). Business Insider. 10 Fantastic Quotes About Markets and Money.

Retrieved from http://www.businessinsider.com/10-great-quotes-about-the-market-

2011-6

Tversky, A., & Kahneman, D. (1971). Belief in the law of small numbers. Psychological

Bulletin, 76(2), 105-110. doi:10.1037/h0031322

62

Venezia, I., Nashikkar, A., & Shapira, Z. (2011). Firm specific and macro herding by

professional and amateur investors and their effects on market volatility. Journal of

Banking and Finance, 35(7), 1599-1609. doi:10.1016/j.jbankfin.2010.11.015

Wilke A. and Mata R. (2012) Cognitive Bias. In: V.S. Ramachandran (ed.) The Encyclopedia

of Human Behavior, vol. 1, pp. 531-535. Academic Press.

William Feather, Brainy Quote (2018). Top 10 William Feather Quotes. Retrieved from

https://www.brainyquote.com/lists/authors/top_10_william_feather_quotes

Zarowin, P., (1989). Does the stock market overrreact to corporate earnings information?.

Journal of Finance 44, 1385-1400.

Documents

REACTIONS OF THE MEXICAN STOCK EXCHANGE AGGREGATE MARKET TO LARGE DAILY PRICE SHOCKS ...catarina.udlap.mx/u_dl_a/tales/documentos/lfp/trueba... · 2018. 7. 18. · price shocks and